LEMLEMWORKING PAPER SERIES

Real Estate and the Great Crisis:Lessons for Macro-Prudential Policy

John V. Duca °Lilit Popoyan *

Susan M. Wachter §

° Research Department, Federal Reserve Bank of Dallas, and Southern

Methodist University, USA* Institute of Economics, Scuola Superiore Sant'Anna, Pisa, Italy

§ Wharton School, University of Pennsylvania, USA

2016/30 July 2016ISSN(ONLINE) 2284-0400

Real Estate and the Great Crisis:

Lessons for Macro-Prudential Policy∗

John V. Duca† Lilit Popoyan ‡ Susan M. Wachter§

Abstract

From a broad macro-financial structure perspective, overly easy credit conditions gave riseto house price booms and busts in several advanced economies (e.g., Ireland, Spain, and theU.S.), and, more specifically in the U.S., an underpricing of risk made possible by regulatoryarbitrage and shadow financing fueled the credit and twin real estate bubbles of the mid-2000s. Across countries and over time bubbles have been particularly acute in real estatemarkets reflecting not only the relatively inelastic supply of land and thin trading of realestate, but also the amplification of shocks via backward-looking price expectations and thefunding of consumption off distorted and elevated prices. The macro-prudential lessons fromthe Great Crisis highlight the need to prevent the build-up of excess real estate financing andlimit the amplification and correlation of real estate risks. And progress has been made ineach of these areas through imposing tougher or new restrictions on the choice sets of lendersor of borrowers, with particulars varying across advanced economies. While regulatory reformof banking is going forward, significant challenges remain especially in dealing with correlatedrisks associated with securitization.

Keywords: Macro-prudential policy, financial crises, credit crunches, real estate bubbles,regulatory arbitrage, risk-taking

JEL classification numbers: G28, E3, R31, R33, R38

∗We thank Michael Weiss and seminar participants at Oxford and the 2016 European Meetings of the Inter-national Finance and Banking Association (IFABs) for comments and suggestions. The views expressed are thoseof the authors and do not necessarily reflect those of the Federal Reserve Bank of Dallas or the Federal ReserveSystem. Lilit Popoyan gratefully acknowledges financial support from European Unions Horizon 2020 grant: No.640772 – Project Dolfins. Any errors are our own.

†Research Department, Federal Reserve Bank of Dallas and Southern Methodist University; E-mail address:john.v.duca@dal.frb.org.

‡Institute of Economics (LEM), Scuola Superiore Sant’Anna, Pisa (Italy); E-mail address:lilitpopoyan@gmail.com.

§Wharton School, University of Pennsylvania; E-mail address: wachter@wharton.upenn.edu

1

1 Introduction

The consensus before the financial crisis of 2007–09 was that monetary policy should avoid

curbing financial excesses if doing so conflicted with the central bank’s goal of keeping inflation

and/or unemployment near target. This view explicitly assumed that monetary policy lacked

enough tools to pursue multiple targets and was too blunt to address financial stability–a view

largely still intact. Implicit was a belief that Basel II made the financial system resilient enough

to survive the unwinding of financial excesses, giving authorities sufficient time to conduct

needed bailouts and for fiscal and monetary policy to clean up damage to the macro-economy and

financial system. In addition, the pre-crisis consensus view also implicitly under-appreciated how

much real estate imbalances could directly contribute to deep recessions and sluggish recoveries.

The 2007–09 financial crisis and recession shattered these implicit assumptions. By mid-2010

when Basel III was announced and when the Dodd-Frank (financial reform) Act (“DFA”) was

passed in the U.S. and the Capital Requirements Directive (“CRD IV”) and Capital Requirement

Regulation (“CRR”) were announced by the EU, the emerging post-crisis consensus accepted

that the financial system was neither resilient enough to survive large financial shocks nor could

policy effectively clean up the macro-economic damage after crises occurred. Systemic risk to

the financial system and real economy was seen as so prevalent, that new macro-prudential

policies were needed to address these risks, allowing monetary policy and fiscal policy to focus

on broad macroeconomic goals (e.g., Blanchard et al. 2010). Although both the theoretical and

empirical framework of macro-prudential policy was still in its infancy–with limited guidance for

policy–the U.S. and European countries, in the wake of the crisis, adopted measures to contain

systemic risks associated with real estate. Given the practical challenges of reaching political

consensus to pass such reforms in the proverbial heat of the moment, there was not enough time

for either Basel III or DFA to be based on a full diagnosis of the financial crisis.

For example, in the U.S., DFA took a kitchen-sink approach of passing widespread guidelines

aimed at curbing financial practices and factors that may have plausibly contributed to the crisis.

The EU, instead, is digesting the regulatory requirements under the CRD IV and CRR that

introduce harmonized prudential rules akin to the Basel III accord. And although much care

and time was taken to write the rules to implement DFA’s guidelines, it was difficult, if not

impossible, for the many regulations to benefit from extensive research on the crisis, much of

which had yet to appear in print. For example, small and midsize banks have complained about

increases in regulatory burden that particularly disadvantage banks that operate at a smaller

scale.

Nine years after the crisis, which began in summer 2007, the financial system continues

working to implement Basel III financial reforms. And in the U.S., eight years after the govern-

ment sponsored entities, Fannie Mae and Freddie Mac were put into conservatorship; reform of

the housing finance system is still in discussion. Over this time, the post-crisis consensus has

continued to evolve in at least two major ways. First, real estate busts are now seen as criti-

cally contributing to deep downturns and sluggish recoveries by impairing the financial system

with losses and households with debt-overhangs. There is recognition that across countries and

over time, bubbles arising in real estate markets can be acute, reflecting not only the relatively

inelastic supply of land and thin trading of real estate, but also the amplification of shocks via

2

backward-looking price expectations and the funding of consumption off distorted and elevated

prices. Second, there is a growing appreciation for the need of macro-prudential policy to address

externalities that cause build-ups of imbalances that lead to crises, directly amplify the effects

of shocks during crises, and indirectly amplify these effects by creating correlated risks. More

recently, there are widespread concerns about whether the long-term growth rates of U.S. and

European GDP have slowed. Are these transitional effects from implementing financial reform

(or residual effects of the unwinding of the crisis) that will eventually unwind or, alternatively,

have financial reforms slowed the underlying dynamism of advanced economies?

Given the depth of the Great Recession and the sluggish recovery, it is appropriate to re-

assess the macro-prudential policy lessons learned, problems solved, and challenges remaining.

We begin by reviewing the lessons that can be drawn from the experience surrounding the

Great Recession, its aftermath, and what financial reforms have done. We then assess how

macro-prudential policy could better address critical risks posed by real estate lending and raise

questions about whether other regulations pose enough macro-prudential benefits to warrant

the burdens imposed. In this endeavor, our approach focuses on the nature of key macro-

prudential externalities–particularly those posed by real estate–and how their manifestation in

real estate valuation swings can be contained. In this way, macro-prudential policy can be better

grounded–both theoretically and empirically–and hopefully, better implemented.

2 Macro-Prudential Lessons Learned From the Great Recession

There is a growing consensus that the bursting of real estate bubbles has contributed greatly

to prolonged downturns (Bordo & Haubrich 2012, Crowe et al. 2013) in being either the impulse

for, (e.g., the Great Recession) or an amplifier of, them (e.g., the Great Depression, Green &

Wachter 2007). In some countries that experienced severe downturns, such as Ireland and Spain,

the demand for real estate was bolstered not only by low mortgage interest rates, but also by

an excessive easing of credit standards via either making too many loans with explicitly high

LTVs (e.g., Ireland–see Kelly et al. 2015) or circumventing covered bond caps on the LTVs of

mortgages by substituting inflated appraised prices for actual transactions prices (as in Spain)

or by avoiding regulatory limits on banks by lending through less regulated depositories (e.g.,

cajas in Spain).1 The resulting high prices triggered building booms that later expanded supply

during the house price bust that ultimately worsened loan losses and crippled financial systems.

Weak regulation took a different form in contributing to the housing bust and the Great

Recession in the U.S., which can be interpreted as being triggered by the collapse of structured

finance that had earlier spawned credit-fueled bubbles in residential and commercial real estate,

the bursting of which ultimately crippled the U.S. financial system and its real economy (Duca

et al. 2010, Levitin & Wachter 2012). An overview of recent evidence on the Great Recession

and shadow banking in the U.S. sheds light not only why real estate downturns are so severe and

1In other countries, higher house prices early in this century owed more to low interest rates fueling upwardshifts in the demand for housing against a relatively price inelastic supply of housing (e.g., the Netherlands orFrance, see Duca et al. 2010). In others, real house prices were flat in the early 2000s reflecting factors such asdeclining population (e.g., Germany and Japan) or tax laws favoring rental housing (e.g., Germany).

3

long-lasting, but also on the tendency for real estate mispricing to give rise to large swings in both

residential and commercial real estate prices and valuations. In addition the U.S. experience

shows how bubbles can arise from the shadow banking sector, an issue still not fully accounted

for by financial reform, as discussed below.

2.1 How and Why Real Estate is Vulnerable to Mispricing

In the U.S., the residential and commercial (“twin”) real estate booms and busts of the

mid-2000s were fueled by the rise and fall of structured finance that funded them. While other

booms and busts in commercial real estate and the real house price boom of the late 1970s

had preceded the twin real estate bubbles of the mid-2000s, those had not been funded on

their upswings by securitization that later dried up. In contrast, the expansion of structured

finance was the key driver of the twin real estate bubbles of the mid-2000s, which coincided

with one another. The last decade’s structured finance boom arose because the tranching in

credit default obligations (CDOs) was made palatable to investors by derivatives enhancements

made possible by the Commodity Futures Modernization Act (CFMA, see Roe 2011, Stout

2011) and because the easing of capital requirements on commercial mortgage-backed (CMBS)

and private-label backed (PMBS) securities provided regulatory arbitrage incentives to fund an

expansion of commercial real estate and nonprime residential lending via securitization. In both

cases, the risk in the underlying real estate investments was underpriced in related–but slightly

different ways–in commercial and residential real estate markets (Levitin et al. 2012).

Using survey data on CRE investors, appraisers, and deal makers, Duca & Ling (2015)

found that the risk premium in investors’ required rate of return on CRE properties was mainly

driven in the short and long run by the combination of a general risk premium (the Baa-10 year

Treasury yield spread) and the effective or marginal capital requirement on CRE loans. The

latter–which affects the liquidity of CRE–reflects the lessor of capital requirements on CRE loans

or CMBS held in bank portfolios and the risk retention for banks that securitize CRE loans.

Duca and Ling attribute the compression of cap rates (akin to a high price-to-earnings ratio) of

the mid-2000s to a decline in the risk premium embedded in investors’ required rate of return

on CRE properties. Their empirical results indicate that the compression of risk premiums

owed to both a reduction in effective capital requirements (the adoption of a 20% risk weight

on high-rated CMBS–see Blundell-Wignall et al. 2008) and a narrowing of the corporate bond

spread (which has a slightly less than one-for-one impact on the long-run risk premium). The

former could be seen as an underpricing of CRE risk by capital regulation. The latter can be

interpreted as an underpricing of CRE risk by investors insofar as they treat the risk on CRE as

akin to that of Baa-corporate bonds that generally are more liquid than commercial properties.

In other words, the implicit use of Baa yields as a benchmark for a required rate of return may

be appropriate for some assets whose returns are based more off value-added (corporate profits

or cash flows) than for CRE whose returns stem from rents and capital gains on land. Levitin

& Wachter (2012) show the trajectory of this underpricing of risk, with risk premia tightening

during the boom.

For PMBS–the main funding source for subprime and Alt-A mortgages–the mispricing of

risk owed to an underassessment of risk that likely reflected a lack of stress testing nonprime

4

mortgages through both business and housing cycles. As Duca et al. (2010) show, while the

limited history of the subprime mortgage delinquency rates before 2007 could be largely tracked

by job growth, this obscured a key role played by house price appreciation. In the early phases

of a financial liberalization, increased credit availability boosts housing demand (see Acolin

et al. 2016, Pavlov & Wachter 2011) and bids up real estate prices. Such gains can delay

the appearance of loan defaults because troubled borrowers could pay off mortgages by selling

or refinancing an appreciated property or continue making loan payments with funds from

home equity loans. This is an acute problem in real estate markets because an underpricing

of default risk in lending–which took the form of low teaser interest rates and excessively easy

credit standards on nonprime mortgages–ultimately leads to inflated asset prices in markets

of fixed supply, as stressed by Pavlov & Wachter (2006, 2009). In addition, at times rising

house prices can paradoxically increase the effective demand for housing because if expected

appreciation is based on extrapolating recent prices (as is observed empirically–see Case et al.

2012), higher prices can raise expected appreciation and lower the perceived real user cost of

housing, thereby fueling further increases in house prices and obscuring underlying loan quality

problems. Moreover, even with the recognition that lax lending terms and underpriced put

options were behind an unsustainable rise in housing prices, selling homes to take advantage of

likely future price depreciation is not feasible (Wachter 2016, Levitin et al. 2012).

Only after the budget constraint effects of higher prices predominated did house prices level

off in 20062 and new subprime borrowers were no longer simply bailed out by price appreciation

(Barakova et al. 2014). Shortly thereafter, subprime delinquencies mounted, inducing investors

to flee nonprime MBS, which induced a reversal of the earlier easing of mortgage credit stan-

dards, setting off a house price bust (Duca et al. 2016). Similar interplay between price and

default dynamics could also delay the appearance of loan quality problems in CRE and mask

the underpricing of risk during the boom phases of real estate bubbles.

In addition, the underpricing of risk was transmitted to the pricing of residential and com-

mercial real estate through somewhat different channels. In commercial real estate (CRE), the

price bubble reflected the transmission of the mispricing of CRE by otherwise savvy investors

through the earnings-multiples that they demanded, whereas the subprime boom reflected the

transmission of underpriced mortgage interest rates as well as unsustainably easy mortgage

credit standards although they both were fueled by regulatory shifts (McCoy et al. 2009).

2.2 How and Why Real Estate Downturns Can Be Prolonged and Slow to

Recover From

There is an emerging consensus that deep and prolonged economic downturns are often linked

to real estate busts both across countries and across time within the U.S. (inter alia, Bordo

& Haubrich 2012). This relationship between real estate bubbles and subsequent economic

crashes has been observed across countries and over time, with the severity and frequency of

2Abraham & Hendershott (1996) describe these dynamics as arising from the tension between the upwardprice momentum effect of appreciation through the real interest rates (“the bubble builder”) versus the negativelong-run tendency for house prices to fall back in line with their fundamentals (“the bubble burster”). The fallback to fundamentals can come quickly when price rises counter the impact of easing borrowing constraints onhomeownership (Barakova et al. 2014) and price expectations reverse.

5

this connection increasing since housing finance has been integrated into global capital markets

(Green & Wachter 2007). There are three underlying channels through which real estate cycles

have recently had a pronounced role in downturns. In addition to traditional or conventional

wealth effects of changes in gross housing wealth, these include changes in the ability to borrow

against housing wealth, mortgage debt overhang effects, and damage to the financial system.

These can be thought of as “real estate financial accelerators,” akin to the more business-oriented

financial accelerator of Bernanke et al. (1996). And each of these are negative externality effects

that lenders do not internalize when making loan decisions, giving rise to more lending than is

socially optimal (Turner 2015, Herring & Wachter 1999).

In recent decades, homeowners have become more able to borrow against accumulated hous-

ing equity gains arising from house price appreciation or paying down mortgage principal. This

phenomenon arose in the UK during the late 1980s and early 1990s–feeding a consumption boom

when house prices rose and a consumption bust when house prices fell (Muellbauer & Murphy

1997)–and has spread to other countries, such as Denmark (Browning et al. 2013), the U.S.

(Hurst & Stafford 2004, Green & Wachter 2007) and Ireland. Micro evidence indicates that this

is mainly a collateral effect, allowing otherwise credit-constrained households to borrow against

housing collateral (inter alia, Hurst & Stafford 2004). The effect operates through the cyclically

changing availability of second liens; while LTVs appeared to be constant through the run-up

to 2007, CLTVs (only knowable after the crisis since real time data were unavailable) increased

dramatically (Levitin & Wachter 2015). The time series evidence for the U.S. indicates that this

ability to extract equity increased over the past several decades (e.g., Carroll et al. 2011), but fell

back during and after the Great Recession (Duca et al. 2013). This up- and then down-swing in

the liquidity of housing wealth amplified the positive and then negative effects of swings in gross

housing assets of U.S. households, and for countries where home equity extraction is feasible

(see Aron et al. 2012, for cross-country evidence).

An additional financial decelerator effect arises from the overhang of mortgage debt accumu-

lated during housing booms that not only impairs housing activity, but also consumption (Mian

& Sufi 2009, 2011). The latter arises because mortgage payment obligations depress consump-

tion either directly by leaving less discretionary income or indirectly by inducing borrowers to

be more cautious in consuming to reduce the risk of incurring future late payment penalties if

the household lacks liquid assets. Collateral price declines exacerbate the latter effect by limit-

ing the ability to refinance existing debt, which must then be repaid as opposed to refinanced.

Empirical studies that disaggregate net worth into debt, gross housing assets, liquid financial

assets, and illiquid or risky financial assets typically find a larger and similar-sized effect of

debt and liquid financial assets on consumption than from the other components of net worth.

Indeed, for the UK, U.S., and Australia, the marginal propensity to consume out of liquid assets

minus debt is roughly five times the size of that of stock wealth and two to three times that

of gross housing assets (see Aron et al. 2012, Duca et al. 2013). Both of these channels, one

operating through the procyclicality of second lien access and the other through more general

procyclicality of mortgage lending standards, affect residential mortgage lending.

The third channel through which a commercial or residential real estate bust has negative

externality effects is through the damage it does to the financial system. By destroying bank

6

capital and causing banks to fall short of capital requirements, the souring of loans extended

during booms induces banks to tighten credit standards (Aron et al. 2012, Bordo et al. 2016)

and curtail lending in general (inter alia, Bernanke & Lown 1991, Kashyap & Stein 2000, Peek &

Rosengren 1995). Indeed, real estate losses–particularly from commercial mortgages (inclusive of

residential construction and land development loans) and CMBS holdings were even more tied to

U.S. bank failures after the onset of the Great Recession than were home mortgages and PMBS

holdings (Antoniades 2015). In addition to reducing financial intermediation, the uncertainty

that real estate busts create about the net wealth of firms and their indirect exposures to

counterparties and customers impairs the needed transparency of risk for financial markets to

operate. The uncertainty about which firms had losses or indirect (e.g., counterparty risk or

off-balance sheet exposures) led to thin trading in which increases in insolvency and liquidity risk

reinforced one another, pushing down both asset prices and trading volumes (see Bernanke 2010)

and pushing up corporate yield spreads. The resulting negative wealth, user cost of capital, and

uncertainty effects of real estate busts on securities markets (direct finance) can thus be viewed

as a third, general type of externality that individual lenders do not fully internalize and which

can result in an above socially optimal build-up of real estate debt in booms. One implication

of the special role played by real estate in financial and economic crises is that the rush to enact

financial reform may have had unintended consequences of over-regulating types of lending that

pose little systemic risk. For example, banks report making fewer small business loans for which

the lack of hard information makes them subject to higher capital assessments in stress tests.

Indeed, some measures show much higher bank regulatory burden (see Bordo et al. 2016, p. 26)

and small business lending has grown slowly in the current U.S. economic recovery which has

been characterized by unusually slow business formation.

3 Progress in Addressing Macro-Prudential Real Estate Risks

Since the financial crisis began, much has been done to detect and address macro-prudential

risks in advanced countries, such as the implementation of Basel III at the national level in

Europe and the implementation and rule-writing specific regulations for the Dodd-Frank Act in

the U.S. The choice of early warning indicators of systemic risk and macro-prudential instru-

ments across these countries reflect similarities and differences in mortgage funding models (see

Wachter 2015, for a comparison between house price booms across these advanced economies).

These can be classified as policies or tools aimed at (a) preventing financial excesses that re-

sult in crises and bolstering resiliency to the financial crises that do occur; and (b) addressing

correlated risks that amplify the broad effects of financial crises. Some of the tools address

more than one of these goals. In addition, each of these goals entails regulations or policies that

are or could be imposed on lenders (“lender-based”), financial markets (“market-based”), and

borrowers/investors (“borrower-based”), and which address the risks posed by innovations that

either circumvent or make obsolete existing regulations, as well as risks posed by networks or

interconnectedness.

7

3.1 Macro-Prudential Tools to Prevent and Limit Financial and Real Estate

Crises

Several major steps have been taken or are being implemented to prevent credit-booms that

can fuel real estate bubbles. These include increased capital requirements and buffers for lenders,

stress tests, limits on risk-taking in securities markets, limits on borrowers, and new liquidity

requirements on banks and money market funds. Some of these steps also limit the intensity

of crises. These include increases in regular and counter-cyclical buffer capital requirements

and the adoption of forward-looking stress tests. In this regard, the counter-cyclical buffer is

particularly useful in reducing the severity of credit crunches. In addition, the new liquidity

requirements on banks and money funds would limit the intensity of runs and of fire sales when

they occur.

3.1.1 Increased Capital Requirements and Buffers for Lenders

First, under Basel III capital requirements were raised on commercial banks in terms of

overall and risk-based leverage ratios, along with the planned implementation of counter-cyclical

capital buffers and additional capital requirements on large and globally systemically important

institutions.3 The specific implementation of these lender-based restrictions has much common-

ality across Europe and the U.S., such as imposing an overall leverage ratio and tightening up the

definition of bank equity capital. In particular, CRD IV in the EU applies generalized counter-

cyclical capital buffer which is conditional on aggregate credit growth and which is also expected

from all member states to limit regulatory arbitrage. Since January 2014, Switzerland–which is

not under EU jurisdiction–has imposed a sector-specific counter-cyclical capital buffer (CCB)

equal to 2% of risk-weighted positions secured by residential property.

Nevertheless, there are some differences in implementation. For example, in Europe, the

counter-cyclical buffer and provisions requiring banks to build buffers are specified in terms of

whether the credit-to-GDP ratio or gap exceeds a certain threshold chosen by individual member

states. Some studies find that such thresholds have empirical evidence favoring their imposition

(e.g., Borio 2014, Drehmann & Juselius 2014, Drehmann & Tsatsaronis 2014) or are effective in

DSGE models (e.g., Alpanda et al. 2014) or in agent-based models (e.g., Popoyan et al. 2015). In

contrast, the U.S. has not (to date) adopted thresholds for adjusting the counter-cyclical buffer

based on an explicit credit-to-GDP ratio. One argument against using such ratios is that data

revisions make them suboptimal on a real-time basis (Edge & Meisenzahl 2011) and another is

that credit outstanding tends to lag business and credit cycles.

Another difference across countries reflects the constrained latitude that Basel III provides

countries with options to impose higher risk weights to some loan categories or capital surcharges

on certain types of riskier loans. For example, in Ireland, Spain and the UK banks are required

to hold additional capital if the share of real estate loans that have high LTVs or DSTIs exceeds

certain thresholds, whereas higher risk weights or capital surcharges are levied on all high LTV

or high DSTI mortgages in Belgium and Switzerland. Indirect sectoral capital requirements that

3In the 2000s large U.S. banks did not build up capital ratios as risks in the financial system rose. And whilelarge U.S. banks built up capital during the 1920s as financial risks mounted in the absence of a large federalsafety net, they needed liquidity and capital support in the Great Depression (see Koch et al. 2016).

8

work on variables that affect capital requirements–such as risk weights (RWs) and loss given

default (LGD) parameters–are more flexibly used under EU jurisdiction (since the counter-

cyclical buffers are generalized and work only on aggregate credit side) due to their targeted

nature. In particular, in Sweden and Norway regulatory authorities imposed a floor on RW

calculated by banks on residential exposures to limit regulatory arbitrage opportunities if a

bank uses the internal rating-based (IRB) model. Perhaps reflecting its pro-home-ownership

or “American Dream” preferences, the U.S. has generally avoided imposing additional capital

surcharges on riskier mortgages for purchasing existing homes. However, the U.S. did impose

a 50 percent capital surcharge on construction and land development loans, partly given the

historical tendency for default rates on these loans to rise more in real estate busts.

3.1.2 Stress Tests

Another major type of bubble-prevention tools aims at buttressing increased capital require-

ments with the imposition of stress tests on banks and systemically important nonbank financial

firms. These tests limit incentives to make riskier loans within broad loan categories and prevent

banks from delaying the write-off of bad loans as a means of complying with capital and safety

requirements. Another advantage of stress tests is that the scenarios incorporated in them can

be used to address forward-looking risks, tail risks, and correlated risks across lenders and the

economy. And stress tests can address the unknown risks of financial innovations by levying

extra capital buffers on the new and “untested” exposures they create. The implementation

of stress tests differs by country, reflecting heterogeneity in the types of shocks or scenarios

they may encounter, which encompasses different judgments about the probability of defaults

and the rates of loss given default for different loans and borrowers operating in different goods

or real estate markets. For example, except for annual EU-wide stress tests conducted by the

European Banking Authority, Norway and Sweden augment their baseline scenarios in their

national stress tests with adverse scenarios for mortgage and real estate markets. As a result,

both countries have tightened the risk weights and default parameters applied to mortgages in

their stress tests.4

From a broader perspective, the widespread adoption of stress tests can be interpreted as

promulgating a lending culture of prudence, where the spirit of the law or regulations–not just the

letter of the law–is followed. In a sense the prudential bank regulation approach of traditionally

more prudent regulators (e.g., Canada and Australia) is now being applied more broadly, albeit

perhaps in a form of more quantitative rather than qualitative judgment from the perspective of

regulators. Nonetheless, the efficacy of stress tests will be limited if potentially large “shadow”

providers of capital escape such tests while they remain in the shadow.

3.1.3 Limits on Risk-Taking in Securities Markets

In addition to adopting tougher capital requirements and stress tests, a third set of bubble-

prevention measures entailed imposing market restrictions on risk-taking in securities markets.

This has not been done so much in Europe, where there are more bank centric financial systems

4Norway tightened assessments of residential mortgage risk, imposing a LGD floor of 20% since January 2014,while Sweden raised the risk weight floor on residential mortgages from 15% to 25% since November 2014.

9

and where efforts have been made to encourage an expanded role of securities markets in pro-

viding finance. Nevertheless, several EU member states have imposed LTV caps on mortgages

eligible to be funded by covered bonds (see ESRB 2014). For example, Austria and Germany

have a 60% LTV limit for residential and commercial mortgages, while Italy, Denmark and Spain

have caps of 80% for residential and 60% for commercial mortgages. In Spain, there is a limit

of 80%, which can be relaxed to 95% if the remaining exposure of the issuing bank is covered

by a guarantee from another bank. In the U.S., LTV caps have not been placed on RMBS, and

more generally private label residential securitization is moribund with RMBS securitized by

the GSEs, Fannie Mae and Freddie Mac, and or by Ginnie Mae for FHA and VA government

guaranteed loans.

In Europe, there are some changes in appraisal rules that affect securitization. Prior to

the crisis, the LTV on a home purchase mortgage could be based on the higher “appraised”

property value rather than the actual transaction price, and LTVs were artificially depressed

to make some mortgages eligible for being financed with covered bonds. Actual sales prices are

now required to calculate LTVs in some countries (e.g., Spain), with national authorities having

the option of imposing larger collateral haircuts to determine eligibility.

In the U.S. where more commercial as well as residential mortgage lending is funded in secu-

rities markets, a major step was taken to impose market-restrictions on risk-taking in securities

markets. Specifically, to limit regulatory arbitrage, the DFA requires originators of CMBS and

of private-label residential MBS that did not meet certain credit standards to retain a first loss

position of 5 percent in such securities originated. This effectively imposes a 5 percent capital

requirements on originators, thereby limiting–but not eliminating–regulatory arbitrage incen-

tives of issuing CMBS and PMBS as a means of funding mortgages. Nevertheless, as Pavlov &

Wachter (2006) show, even with modest skin in the game, today’s fees on mortgage lending will

still be attractive if the “put option” is in the money and NPV of mortgage lending is expected

to be negative.

Another step taken in the U.S. concerns new limits on which kinds of home mortgages can

be securitized. Now such mortgages are limited to either government guaranteed mortgages

(FHA and VA), mortgages meeting the underwriting criteria of Fannie Mae or Freddie Mac, or

mortgages that have DSTI ratios below 43 percent with other restrictions on borrower attributes.

However, this remains an open issue as the resolution of GSE reform is uncertain. The CMBS

market has had a limited resurgence amid these requirements, although as noted, PMBS has

not.

In straight-forward ways, the imposition of stress tests, new capital requirements, and risk

retention rules reduce the risk that financial excesses will build up and thereby reduce the risk

that financial crises occur. Nevertheless, there are several ways that market limits on risk-

taking could be further improved. One, in particular, involves the cost of securitization relative

to holding loans on portfolio. The initial round of the DFA increased the capital requirements on

loans held in bank portfolios by about 2.5 percentage points somewhat less than they increased

the effective capital requirements by about 5 percentage points on securitizing assets through risk

retention rules. This had the effect of directly improving the safety of banks while reducing some

of the regulatory arbitrage incentives to circumvent bank capital requirements via securitization.

10

However, there may be increased regulatory arbitrage via securitization unless the upcoming

imposition of additional capital requirements in the form of counter-cyclical capital buffers and

capital surcharges on SIFIs is matched by a further increase in the risk retention rules. As a

result, any “direct” increase in the safety of banks from the phase-in of additional capital rules

could be offset by increased securitization–especially in commercial real estate which is less

subject to new regulations imposed on all originators than in the case of home mortgage lending

and potentially in residential as well, if Fannie Mae and Freddie Mac are phased out. In addition,

in the absence of a concomitant increase in risk retention rules, unless downpayment and debt-

service requirements are imposed on nonconventional residential mortgages, the 5 percent risk

retention rules may not adequately prevent a socially excessive level of nonprime mortgage

origination funded, for example, by hedge funds which may not be treated as systemically

important.

3.1.4 Limits on Borrowers

Another set of initiatives to limit the build-up of systemic risk involve direct restrictions on

borrowers’ ability to borrow (these differ from greater bank capital requirements on risky loans

mentioned above). These have differed across countries, in part reflecting a lack of theoretical

consensus about their efficacy. For example in DSGE models with endogenous distortions, re-

ducing maximum limits on individual borrowing (e.g., lowering LTV caps) in periods of rapid

credit growth or high debt-to-GDP ratios can improve financial stability (Brunnermeier & San-

nikov 2012, Rubio & Carrasco-Gallego 2014, Lambertini et al. 2013). Other DSGE models point

to the need for macro-prudential tools only with endogenous distortions; without such distor-

tions, these models imply monetary policy should be used to quell credit bubbles rather than

macro-prudential tools (e.g., Cecchetti & Kohler 2012). In agent-based models, such restrictions

promote both financial and macroeconomic stability (e.g., Popoyan et al. 2015).

The empirical evidence generally indicates a financial stability role for tools limiting or dis-

couraging lending to high risk borrowers, but less about which specific tools are most efficacious.

Cross-country studies generally find that limits on borrowers such as caps on LTVs, debt service-

to-income (DSTI) or debt-to-income (DTI) ratios are all effective in limiting large buildups in

debt relative to income (e.g., Borio & Shim 2007, Lim et al. 2011, Ciani et al. 2014, Dell’Ariccia

et al. 2012). Kuttner & Shim (2013), however, find evidence that limits on borrower debt

service-to-income ratios tend to be more effective than LTV or DTI caps in stabilizing housing

credit growth and house price appreciation. Interestingly, the U.S. has eschewed imposing LTV

caps on mortgages not insured by federal agencies, in favor of shielding lenders from borrower

lawsuits on mortgages meeting DSTI caps and other non LTV criteria.

In contrast to the U.S., Asian and European countries have been more open to imposing

LTV caps across borrowers and to varying them counter-cyclically. For example, Hong Kong

has imposed LTV caps, which it has adjusted to counter overvaluation in real estate. In the

wake of the financial crisis, several European countries have imposed general LTV caps (e.g.

Sweden, Belgium and the Netherlands) or more stringent ones on loans for high priced homes

to first-time home-buyers or on investors (“buy-to-let”). While Canada does not have explicit

LTV caps in general, its government insurer of mortgages has varied its LTV and DSTI limits

11

on the mortgages it insures, while the regulatory authority reduced the maximum maturity of

mortgages from 30 to 25 years to counter rising house price pressures from low interest rates

(see Crawford 2015).

3.1.5 New Liquidity Requirements on Banks and Money Market Funds

A final set of actions to help prevent financial crises and limit their severity are several

new liquidity requirements. The two most prominent of these are aimed at ensuring banks

have sufficient liquidity to withstand stressful situations. One is the Liquidity Coverage Ratio

(LCR), which requires banks to hold enough high quality liquid assets (HQLA) that could be

sold to offset withdrawals of short-term funds over a 30-day period. The second is the net stable

funding ratio (NSFR), which requires banks to have enough stable liabilities to fund assets

over the next year under different stress scenarios. The above new liquidity requirements were

generally applied and implemented similarly across the U.S. and European countries with some

minor exceptions.5

However, reflecting the prevalence of floating net asset value (NAV) money market mutual

funds in Europe and the former prevalence of fixed-NAV money funds in the U.S., several steps

were taken in the U.S. (but not Europe) to improve the liquidity of money funds to limit the

risk of runs and the fire sales of commercial paper. Most notable of these is the requirement

that institutional money funds investing in non-Treasury paper use floating net asset values to

price accounts. Earlier, U.S. money funds had a “don’t break the buck” policy of not pricing

money fund accounts below par when the assets the funds invested in fell in price. When large

losses on commercial paper emerged around Lehman Brother’s failure in September 2008, many

account holders withdrew at par rather than risk capital losses if their fund failed. Investors in

one major fund that did fail suffered capital losses, which though small in absolute size, were

large relative to the narrow returns typically earned on short-run paper. This helps account

for the 1

2trillion dollar run on U.S. institutional money funds in a three week period following

Lehman’s collapse.

The new liquidity requirements on banks and money funds reduce the risk of financial failure

in three ways. First, they lower the risk of funding drying up on banks which could otherwise

be tempted to excessively rely on uninsured funds and on institutional money funds whose prior

“don’t break the buck” account pricing could induce runs. This lowers the risk that a drying up

of funding could induce a credit crunch at banks or a decline in the demand to hold commercial

paper by money funds. Second, this lowers the risk of fire sales of private paper or bonds by

banks and institutional money funds, which can feedback on asset prices and amplify a financial

shock into a full blown financial crisis.6 Third, the liquidity requirements raise the financial

system’s costs of funding risky, longer duration assets with short-run debt that would otherwise

have the externality effect of underpricing the tail risk of a financial crisis. In particular, investors

in short-term debt may think that the risk they take is limited because they have the ability

5For example, new Congressional legislation mandates that U.S. municipal bonds be treated as high qualityliquid assets despite the recent bankruptcies of Detroit and Puerto Rico.

6 Duca (2013a,b) notes that the commercial paper market fully melted down in the Great Depression andthat, thanks in part to federal interventions, a full meltdown–but not a partial meltdown–was avoided in theGreat Recession.

12

to sell such short-term debt should the risk of the debt issuer rise. However, in the event of

a tail risk–such as a financial crisis where private paper in general becomes less liquid–there is

little ability to limit risk by investing in short-term debt that rolls over rather than in long-term

debt. As a result, the true funding cost of financing risk assets is underpriced because tail risk

externalities are underpriced, thereby encouraging overinvestment in risky, illiquid assets such

as real estate.

The new liquidity requirements help address all three effects by limiting the incentives of

banks and money funds to use short-run funds to finance investments in longr duration risky

assets (“carry trade” behaviour)–such as CMBS or PMBS–or to invest in the short-run debt of

other financial institutions used to finance investments in risky assets. One issue is whether the

new rules on money funds go far enough. A prime example is that retail money funds are not

required to adopt floating NAV pricing, which not only continues posing a risk of runs on retail

funds, but also encourages some institutional funds to convert to retail status by limiting the

size of individual money fund accounts. While the latter action may not fully offset the rule,

institutional investors can circumvent the floating NAV rules on institutional money by breaking

their former large-sized accounts into many smaller-sized accounts. Such actions consequently

limit the effectiveness of floating NAV pricing rules on institutional funds to reduce the risk of

money fund runs.

3.1.6 Market-Based Backstop Facilities

Another set of macro-prudential tools that reduce the intensity of crises and aide recovery

from them are the market-based backstop facilities created by the Federal Reserve and Treasury,

referred to as lender-of-last-resort or credit-easing tools. Although these tools may be seen as

preserving the functionality of particular markets, there are sizable macroeconomic externalities

from possible market dysfunctionality that impart a macro-prudential dimension to these tools.

In response to the financial crisis, there was concern that Federal Reserve support of par-

ticular institutions under earlier “13-3(c)” powers encouraged too much moral hazard by large

institutions. DFA circumscribes lender of last resort actions insofar as the central bank can

set up backstop facilities for a broad array of market participants to use with the approval of

the U.S. Treasury provided that such facilities are not designed to save any one or handful of

institutions. During the financial crisis of 2007–09, five major actions (among more that space

prescribes mentioning) were taken by the Fed to backstop markets which continue to be options

allowed under DFA. One such facility was the Asset-Backed Commercial Paper Money Market

Mutual Fund Liquidity Facility AMLF, under which the Federal Reserve made nonrecourse loans

to money funds investing in private commercial paper. This too had the effect of cushioning

the negative impact of the financial and money market crisis on commercial paper issuance

(see Duygan-Bump et al. 2013). Another facility was the Commercial Paper Funding Facility

(CPFF), under which the Federal Reserve bought A1/P1 rated commercial paper at interest

rates that were 100 basis points above the option-indexed swap rate. At one point, the Federal

Reserve held about 20 percent of U.S. commercial paper outstanding and its actions prevented

commercial paper from collapsing as much as it did during the Great Depression (Duca 2013a).

Both of these actions were augmented by the U.S. Treasury’s action to extend deposit insurance

13

to money fund accounts and to insure many large time deposits issued by banks.

In addition to buttressing commercial paper and money funds, the Federal Reserve created

three other noteworthy facilities (among many) to prevent liquidity problems from worsening the

crisis. These included creating the Term Auction Facility (TAF) to provide banks with longer

term discount loans and through a new facility with less stigma, the primary dealer facility to

provide liquidity backstop to dealers who make markets in initially offered debt and equity, and

creating facilities to fund eligible consumer, small business, and high-grade CMBS. Of these,

the TAF helped make banks more liquid, allowing them to continue lending to and servicing

real estate markets, while the last facility helped the top quality portion of the secondary

CMBS market to continue operating–setting the stage for the later reemergence of limited CMBS

issuance.

3.2 Tools to Address Risks Posed by Network Externalities

Aside from the macroeconomic effects that crises create by directly impairing lenders and

the functioning of securities markets,7 there are important network risks that macro-prudential

tools should address. For real estate assets whose collateral role can obscure tail risk, it is critical

to monitor lending from all sources that might unsustainably elevate collateral prices. LTV caps

and tougher capital requirements are ineffective if values are elevated by other sources of capital.

A prominent remaining correlated risk concerns the impact on consumption of mortgage

equity withdrawals, which tend to boost consumption in liberalized economies during booms and

restrain it when debt overhang effects prevail in their aftermath (see Aron et al. 2012, Muellbauer

& Murphy 1997). Most of the active mortgage equity withdrawal during the subprime housing

boom took the form of cash-out mortgage refinancing or borrowing via home equity lines of

credit. Starting in late 2007, the incidence of the former has been curtailed by Fannie Mae and

Freddie Mac, which have imposed sizable surcharge fees withdrawing equity when refinancing

mortgages. And new regulatory and enforcement scrutiny by the Consumer Finance Protection

Bureau (CFPB) coupled with stress tests by other bank regulators have likely deterred riskier

forms of mortgage equity withdrawal.

More generally, unless new econometric techniques are employed (Duca et al. 2013, being an

exception), it is difficult to track how much the liquidity of housing wealth has been affected by

new regulatory actions, thereby limiting the impact of house prices on consumer spending. Fur-

thermore and moreover, without an explicit macro-prudential goal of limiting mortgage equity

withdrawal effects, there is no guarantee that the new cash-out surcharges by Fannie Mae and

Freddie Mac will continue or might be offset by more liberal policies by private MBS securitizers

or by banks holding home equity loans in portfolio. Unlike other assets, holders of first liens

are not required to be notified or give their permission for granting second liens; in part due

to this, the capacity to track property level combined loan-to-value ratios (CLTVs) is limited

(Levitin & Wachter 2015). Thus, this important source of risk remains. Indeed, as documented

by Kumar (2015), the 80 percent cap that Texas imposed on CLTVs for loans extracting home

equity significantly limited the rise of home equity mortgage delinquencies compared to that of

7These include credit crunches, higher debt costs, and stock or bond wealth effects on consumer spending andon the ability of firms and consumers to raise funding.

14

other U.S. states during the recent housing bust.

Other correlated risks concern how problems in one market can cascade into others, particu-

larly if there are complex interactions and exposures. Aside from the fire sale issue, derivatives

pose serious network risk exposures as was revealed in the financial crisis when CDS contracts

suddenly became suspect as the contracts were not transparent and uniform, creating uncer-

tainty about the exposures of private firms and impairing securities markets in general (Bernanke

2008). Reforms to address this include creating industry standards and a clearinghouse for many

derivatives contracts, spurred in part by the experience in 2008 surrounding the real estate ex-

posures of Bear Stearns and Lehman.8 Nevertheless, there are concerns that these may be

insufficient to adequately prevent crises, such as that of 2008 (Duffie & Zhu 2011).

Another macro-prudential tool for gauging both network effects and markets where imbal-

ances are building are heat or radar maps designed to summarize and provide an overview of

risks in the financial system–the development of which was recommended by Geanakoplos (2009,

2010). One example is the “heat map” of Aikman et al. (2015), which highlights that CRE val-

uations are a major outlying risk, consistent with the cap rate analysis of Duca & Ling (2015).

Another type maps out networks such as that of Gai et al. (2011).

Nevertheless, these are tools for identifying systemic risks, not necessarily for preventing or

addressing them, which requires additional action. The development of such tools has enabled

regulators to identify and act upon areas of risk, such as commercial real estate as noted by

Federal Reserve Chair Yellen and the Board of Governors of the Federal Reserve System (2015).

One particular issue for such tools is the danger of rapid growth in new products not on

regulators’ radar maps. A poignant example is how in the lead up to the subprime crisis, the

only widely available measures of residential loan-to-value ratios were based on conventional

loans and for all home-buyers and indicated no problem, whereas gauges on LTVs for first-

time homebuyers who had access to nonprime mortgages showed a sharp deterioration in credit

standards (Duca et al. 2011). More broadly, radar or heat maps will need to be capable of

tracking a rapid expansion of less-regulated financial sectors (Wachter 2016).

Indeed, the empirical evidence indicates that the shadow banking system’s provision of short-

term business credit (Duca 2016) and of commercial real estate finance (Levitin & Wachter 2013,

Duca & Ling 2015) is prone to expanding rapidly when general risk aversion recedes or when

regulatory arbitrage increases because of either financial innovation or regulatory laxity. And

since shadow or security market financed lending depends on uninsured funding, these credit

sources have been very vulnerable to runs and credit crunches during periods of distress. Owing

to the volatility and potential size of shadow funding, the thin-trading of real estate and the

lagging adjustment of construction, swings in shadow bank funding of real estate can induce

real estate cycles that can undermine financial and macroeconomic stability even if banks are

well-regulated. For this reason, tightening regulations on banks needs to be matched by tougher

regulations on other sources of funding (such as toughening risk retention requirements on loan

8As Duffie & Zhu (2011) note, “Effective clearing mitigates systemic risk by lowering the likelihood thatdefaults propagate from counterparty to counterparty. Clearing could also reduce the degree to which the solvencyproblems of a market participant are suddenly compounded by a flight of its OTC derivative counterparties suchas when the solvency of Bear Stearns and Lehman Brothers was in question. Central clearing reduces the risk ofdisruptions to financial markets through fire sales of derivatives positions or of collateral held against derivativespositions.”

15

securitizers) to address the externalities those sources pose.9

4 Concluding Comments

From a broad macro-financial structure perspective, an underpricing of risk made possible

by regulatory flaws contributed greatly to the housing and financial crises of the late-2000s in

several advanced economies. In the U.S. such shortcomings enabled regulatory arbitrage and

shadow financing to fuel the credit and twin real estate bubbles of the mid-2000s that led to

the Great Recession. Insufficient macro-prudential policy also contributed to large real estate

booms and busts in Ireland and Spain, whose economies were among the hardest hit in Europe

before the sovereign debt crisis began in the spring of 2010. The effects of regulatory failure

and mispricing of risk were most acute in real estate markets reflecting not only the relatively

inelastic supply of land and thin trading of real estate, but also the amplification of shocks via

backward-looking price expectations and the funding of consumption off distorted and elevated

prices. The reversal of these excesses triggered downturns amplified by correlated shocks to

financial intermediation and–in the cases of the U.S. and Ireland–to consumption. The resulting

debt and residential real estate overhangs both deepened the Great Recession and prolonged the

subsequent sluggish recovery from all three of these countries.

On the surface, this pattern of macroeconomic and financial behavior lends itself to inter-

preting the Great Recession as the outcome of a largely endogenous leverage or debt-super cycle.

That interpretation, however, does not really account for the 80 year hiatus between the Great

Depression and Great Recession in the U.S. Nor does it account for why twin booms and busts

emerged in both commercial and residential real estate prices in the U.S. or for the rise and fall

of structured finance and shadow banking that accompanied them during the past two decades.

More detailed analysis indicates that changes in regulation amplified financial innovation

in the form of structured finance largely as part of the shadow bank system. This, in turn,

spawned collateralized credit bubbles and fueled twin real estate bubbles and debt overhangs

that ultimately crippled the U.S. financial system and its real economy. From this perspective,

the leverage cycle of this century’s first decade was not simply an endogenous development (e.g.,

Minsky 1982a,b). Rather regulatory actions undermined the restraining influence of capital

requirements on credit creation, thereby spiking the proverbial punch bowl instead of taking it

away. Hence better regulatory steps can help tame or temper the leverage cycle (Geanakoplos

2009, 2010).

The macro-prudential lessons from the Great Crisis highlight the need to prevent the build-

up of excess real estate financing, limit the amplification and correlation of real estate risks, and

enhance the ability to clean up real estate busts. More progress has been made in limiting the

build-up of excess home purchase lending than in restricting excess price pressures in commercial

9For example, following the stock market crash of 1987, Federal Reserve was able to prevent a full-blownfinancial and economic crisis partly because the crash reflected an equity bubble that was not financed by debtand also because by providing liquidity to banks and indirectly to securities dealers, the Fed could stabilize afinancial system in which only one-third of household and nonfinancial business debt was funded by securitiesmarkets. In contrast, the financial imbalances leading up to the Great Recession were debt-financed and abouttwo-thirds of household and nonfinancial business debt was funded by securities markets when the subprime crisishit. As a result, only regulating bank balance sheets and providing liquidity to just banks were inadequate toaddress the recent crisis.

16

real estate. With respect to securities markets and the shadow bank system in the U.S., progress

has also been made in reducing the risk of fire sale actions by banks and money funds than

by other investors in CMBS. At a more concrete level, most of this has occurred by limiting

debt-service burdens on households, implementing limited skin-in-the-game risk retention on

parties creating mortgage-backed securities, instituting stress tests on banks and “systemically

important financial institutions,” and imposing liquidity limits on commercial banks and money

funds.

There, however, remain major challenges. For one, some of the increased regulation of small

business lending and community banks may have had the unintended consequence of stifling

business formation and the economic recovery while doing little to address systemic risks that

are more centered on real estate and high-leverage lending. Here, some well-targeted and narrow

deregulatory steps may be warranted, but not a weakening of capital buffers or risk retention

rules for securitization. Second, restricting the extent of home equity borrowing via loan-to-

value caps could limit the build-up of mortgage debt overhangs and their negative consequences

as illustrated by recent experience in Texas (see Kumar 2015); even tracking the extent to which

home equity lending is increasing a property’s CLTV remains difficult (see Levitin & Wachter

2015). Third, while general caps or Canadian-style limits on LTVs for home purchase mortgages

could limit the risks posed by procyclical lending by shadow banks, the U.S. has so far decided

not to impose general LTV limits unlike other countries (see Crowe et al. 2013). Instead, the

U.S. has favored limits on debt service-to-income ratios, which international evidence may now

favor (Kuttner & Shim 2013). Fourth, the relatively less-increased regulation of lending to the

commercial real estate sector relative to home real estate sector may help account for why recent

CRE valuations are notably more elevated than house price valuations, induced in large part by

low long-term real interest rates. Fifth, the Federal Reserve’s recent announcement of higher

capital requirements on systemically important financial institutions effectiveness in curbing real

estate risks could be undermined by regulatory capital arbitrage unless there is a strengthening

of “skin-in-the-game” risk retention rules for loss provisions on mortgage securitizers or of capital

requirements on all holders of nonqualified mortgages. This may also be an issue for European

nations that increase capital requirements on their systemically important institutions. Finally,

there are uncertainties about the future structure of RMBS (e.g., GSE reform in the U.S.) and

what entities would be responsible for (and hopefully effective in) maintaining underwriting

standards. For these reasons, imposing macro-prudential limits on home equity borrowing could

deter homeowners from stepping under debt-overhangs, while strengthening effective capital

requirements and other regulations on securitized CRE and home mortgages could help keep

real estate markets from reentering the shadows of structured finance.

References

Abraham, J. M. & Hendershott, P. H. (1996), ‘Bubbles in metropolitan housing markets’, Journal ofHousing Research 7, 191–207.

Acolin, A., Bricker, J., Calem, P. S. &Wachter, S. M. (2016), ‘Borrowing constraints and homeownership’,American Economic Review 106, 625–29.

17

Aikman, D., Kiley, M. T., Lee, S. J., Palumbo, M. & Warusawitharana, M. (2015), ‘Mapping heat in theU.S. financial system’, Finance and Economics Discussion Series (2015-059).

Alpanda, S., Cateau, G. & Meh, C. (2014), ‘A policy model to analyze macroprudential regulations andmonetary policy’, BIS Working Paper (461).

Antoniades, A. (2015), ‘Commercial bank failures during the great recession: the real (estate) story’,ECB Working Paper (1779).

Aron, J., Duca, J. V., Muellbauer, J., Murata, K. & Murphy, A. (2012), ‘Credit, housing collateral, andconsumption: evidence from Japan, the UK, and the U.S.’, Review of Income and Wealth 58(3), 397–423.

Barakova, I., Calem, P. S. & Wachter, S. M. (2014), ‘Borrowing constraints during the housing bubble’,Journal of Housing Economics 24, 4–20.

Bernanke, B. (2010), ‘Implications of the financial crisis for economics’, speech to Conference Co-sponsored

by the Center for Economic Policy Studies and the Bendheim Center for Finance, September

24(Princeton University).

Bernanke, B., Gertler, M. & Gilchrist, S. (1996), ‘The financial accelerator and the flight to quality’,Review of Economics and Statistics 78, 1–15.

Bernanke, B. S. (2008), Financial regulation and financial stability, in ‘Speech to the the Federal De-posit Insurance Corporations Forum on Mortgage Lending for Low and Moderate Income Households,Arlington, VA, July’, Vol. 8.

Bernanke, B. S. & Lown, C. S. (1991), ‘The credit crunch’, Brookings Papers on Economic Activity

1991(2), 205–247.

Blanchard, O., Dell’Ariccia, G. & Mauro, P. (2010), ‘Rethinking macroeconomic policy’, Journal of

Money, Credit and Banking 42(s1), 199–215.

Blundell-Wignall, A., Atkinson, P. et al. (2008), ‘The subprime crisis: Causal distortions and regulatoryreform’, Lessons from the Financial Turmoil of 2007 and 2008 (Reserve Bank of Australia).

Board of Governors of the Federal Reserve System (2015), ‘Monetary report to the congress’, July .

Bordo, M. D., Duca, J. V. & Koch, C. (2016), ‘Economic policy uncertainty and the credit channel:Aggregate and bank level us evidence over several decades’, NBER Working Paper (22021).

Bordo, M. D. & Haubrich, J. G. (2012), ‘Deep recessions, fast recoveries, and financial crises: Evidencefrom the american record’, NBER Working Paper (18194).

Borio, C. (2014), ‘The financial cycle and macroeconomics: What have we learnt?’, Journal of Banking& Finance 45, 182–198.

Borio, C. E. & Shim, I. (2007), ‘What can (macro-) prudential policy do to support monetary policy?’,BIS Working Paper (242).

Browning, M., Gørtz, M. & Leth-Petersen, S. (2013), ‘Housing wealth and consumption: a micro panelstudy’, The Economic Journal 123(568), 401–428.

Brunnermeier, M. K. & Sannikov, Y. (2012), ‘A macroeconomic model with a financial sector’, American

Economic Review 104(2), 379–421.

Carroll, C. D., Otsuka, M. & Slacalek, J. (2011), ‘How large are housing and financial wealth effects? anew approach’, Journal of Money, Credit and Banking 43(1), 55–79.

Case, K. E., Shiller, R. J. & Thompson, A. (2012), ‘What have they been thinking? home buyer behaviorin hot and cold markets’, NBER Working Papers (18400).

Cecchetti, S. G. & Kohler, M. (2012), ‘When capital adequacy and interest rate policy are substitutes(and when they are not)’, BIS Working Paper (379).

18

Ciani, D., Cornacchia, W., Garofalo, P. et al. (2014), ‘The macroprudential measures adopted in Europefor the real estate sector’, Bank of Italy Occasional Papers (227).

Crawford, A. (2015), ‘Building stable mortgage markets: Lessons from canada’s experience’, Journal ofMoney, Credit and Banking 47(S1), 81–86.

Crowe, C., Dell’Ariccia, G., Igan, D. & Rabanal, P. (2013), ‘How to deal with real estate booms: Lessonsfrom country experiences’, Journal of Financial Stability 9(3), 300–319.

Dell’Ariccia, G., Igan, D., Laeven, L. & Tong, H. (2012), ‘Policies for macrofinancial stability: How todeal with credit booms’, IMF Staff Discussion Note (12/06).

Drehmann, M. & Juselius, M. (2014), ‘Evaluating early warning indicators of banking crises: Satisfyingpolicy requirements’, International Journal of Forecasting 30(3), 759–780.

Drehmann, M. & Tsatsaronis, K. (2014), ‘The credit-to-gdp gap and countercyclical capital buffers:questions and answers’, BIS Quarterly Review, March .

Duca, J. V. (2013a), ‘Did the commercial paper funding facility prevent a great depression style moneymarket meltdown?’, Journal of Financial Stability 9(4), 747–758.

Duca, J. V. (2013b), ‘The money market meltdown of the great depression’, Journal of Money, Credit

and Banking 45(2-3), 493–504.

Duca, J. V. (2016), ‘How capital regulation and other factors drive the role of shadow banking in fundingshort-term business credit’, Journal of Banking & Finance 69(S1), S10–S24.

Duca, J. V. & Ling, D. C. (2015), ‘The other (commercial) real estate boom and bust: the effects of riskpremia and regulatory capital arbitrage’, manuscript .

Duca, J. V., Muellbauer, J. & Murphy, A. (2010), ‘Housing markets and the financial crisis of 2007–2009:lessons for the future’, Journal of Financial Stability 6(4), 203–217.

Duca, J. V., Muellbauer, J. & Murphy, A. (2011), ‘House prices and credit constraints: Making sense ofthe US experience’, The Economic Journal 121(552), 533–551.

Duca, J. V., Muellbauer, J. & Murphy, A. (2013), ‘How financial innovations and accelerators drivebooms and busts in us consumption’, Oxford University (mimeo).

Duca, J. V., Muellbauer, J. & Murphy, A. (2016), ‘How mortgage finance reform could affect housing’,American Economic Review 106, 620–24.

Duffie, D. & Zhu, H. (2011), ‘Does a central clearing counterparty reduce counterparty risk?’, Review of

Asset Pricing Studies 1(1), 74–95.

Duygan-Bump, B., Parkinson, P., Rosengren, E., Suarez, G. A. & Willen, P. (2013), ‘How effective werethe federal reserve emergency liquidity facilities? Evidence from the asset-backed commercial papermoney market mutual fund liquidity facility’, The Journal of Finance 68(2), 715–737.

Edge, R. M. & Meisenzahl, R. R. (2011), ‘The unreliability of credit-to-gdp ratio gaps in real-time:Implications for countercyclical capital buffers’, International Journal of Central Banking 7(4), 261–298.

ESRB (2014), ‘The esrb handbook on operationalising macro-prudential policy in the banking sector’,March .

Gai, P., Haldane, A. & Kapadia, S. (2011), ‘Complexity, concentration and contagion’, Journal of Mon-

etary Economics 58(5), 453–470.

Geanakoplos, J. (2009), The leverage cycle, in ‘NBER Macroeconomics Annual 2009, Volume 24’, Uni-versity of Chicago Press, pp. 1–65.

Geanakoplos, J. (2010), ‘Solving the present crisis and managing the leverage cycle’, New York Federal

Reserve Economic Policy Review 16(1), 101–131.

19

Green, R. K. & Wachter, S. M. (2007), ‘The housing finance revolution’, Proceedings-Economic Policy

Symposium-Jackson Hole pp. 21–67.

Herring, R. J. & Wachter, S. M. (1999), ‘Real estate booms and banking busts: an international perspec-tive’, The Wharton School Research Paper (99-27).

Hurst, E. & Stafford, F. (2004), ‘Home is where the equity is: Mortgage refinancing and householdconsumption’, Journal of Money, Credit and Banking 36(6), 985–1014.

Kashyap, A. K. & Stein, J. C. (2000), ‘What do a million observations on banks say about the transmissionof monetary policy?’, American Economic Review 90, 407–428.

Kelly, R., McCannb, F. & OTooleb, C. (2015), ‘Credit conditions, macroprudential policy and houseprices’, Central Bank of Ireland Research Technical Paper (06RT15).

Koch, C., Richardson, G. & Van Horn, P. (2016), ‘Bank leverage and regulatory regimes: Evidence fromthe great depression and great recession’, The American Economic Review 106(5), 538–542.

Kumar, A. (2015), ‘Do restrictions on home equity extraction contribute to lower mortgage defaults?evidence from a policy discontinuity at the Texas’ border’, Federal Reserve Bank of Dallas Working

Paper (1410).

Kuttner, K. N. & Shim, I. (2013), ‘Can non-interest rate policies stabilize housing markets? evidencefrom a panel of 57 economies’, NBER Working Paper (19723).

Lambertini, L., Mendicino, C. & Punzi, M. T. (2013), ‘Leaning against boom–bust cycles in credit andhousing prices’, Journal of Economic Dynamics and Control 37(8), 1500–1522.

Levitin, A. J., Pavlov, A. D. & Wachter, S. M. (2012), ‘Dodd-frank act and housing finance: Can itrestore private risk capital to the securitization market, the’, Yale Journal on Regulation 29, 155–180.

Levitin, A. J. & Wachter, S. M. (2012), ‘Explaining the housing bubble’, Georgetown Law Journal

100(4), 1177–1258.

Levitin, A. J. & Wachter, S. M. (2013), ‘The commercial real estate bubble’, Harvard Business Law

Review (3), 83–118.

Levitin, A. J. & Wachter, S. M. (2015), ‘Second liens and the leverage option’, Vanderbilt Law Review

68, 1243–1294.

Lim, C. H., Costa, A., Columba, F., Kongsamut, P., Otani, A., Saiyid, M., Wezel, T. & Wu, X. (2011),‘Macroprudential policy: what instruments and how to use them? lessons from country experiences’,IMF Working Paper (11/238).

McCoy, P. A., Pavlov, A. D. & Wachter, S. M. (2009), ‘Systemic risk through securitization: The resultof deregulation and regulatory failure’, Connecticut Law Review 41, 493–541.

Mian, A. & Sufi, A. (2009), ‘The consequences of mortgage credit expansion: Evidence from the USmortgage default crisis.’, Quarterly Journal of Economics 124(4), 1449–1496.

Mian, A. & Sufi, A. (2011), ‘House prices, home equity–based borrowing, and the US household leveragecrisis’, The American Economic Review 101(5), 2132–2156.

Minsky, H. P. (1982a), ‘Can it happen again? essay on instability and finance’, New York: ME Sharpe .

Minsky, H. P. (1982b), ‘The financial instability hypothesis: capitalist processes and the behavior of theeconomy, 1979’, Financial crisis: theory, history and policy, Cambridge University Press & Maison

des Sciences de lHomme pp. 13–39.

Muellbauer, J. & Murphy, A. (1997), ‘Booms and busts in the U.K. housing market’, The Economic

Journal 107(445), 1701–1727.

Pavlov, A. D. & Wachter, S. M. (2009), ‘Systemic risk and market institutions’, Yale Journal on Regu-

lation 26(2), 445–455.

20

Pavlov, A. & Wachter, S. (2011), ‘Subprime lending and real estate prices’, Real Estate Economics

39(1), 1–17.

Pavlov, A. & Wachter, S. M. (2006), ‘The inevitability of marketwide underpricing of mortgage defaultrisk’, Real Estate Economics 34(4), 479–496.

Peek, J. & Rosengren, E. (1995), ‘The capital crunch: neither a borrower nor a lender be’, Journal ofMoney, Credit and Banking 27(3), 625–638.

Popoyan, L., Napoletano, M. & Roventini, A. (2015), ‘Taming macroeconomic instability: Monetary andmacro prudential policy interactions in an agent-based model’, LEM Working Paper Series (2015/33).

Roe, M. J. (2011), ‘The derivatives market’s payment priorities as financial crisis accelerator’, StanfordLaw Review 63, 539.

Rubio, M. & Carrasco-Gallego, J. A. (2014), ‘Macroprudential and monetary policies: Implications forfinancial stability and welfare’, Journal of Banking & Finance 49, 326–336.

Stout, L. A. (2011), ‘Derivatives and the legal origin of the 2008 credit crisis’, Harvard Business Law

Review 1, 1–38.

Turner, A. (2015), Between Debt and the Devil: Money, Credit, and Fixing Global Finance, PrincetonUniversity Press.

Wachter, S. (2015), ‘The housing and credit bubbles in the United States and Europe: A comparison’,Journal of Money, Credit and Banking 47(S1), 37–42.

Wachter, S. M. (2016), Informed Securitization, University of Pennsylvania Press.

21